New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Slightly different results for SEQ and PAR #335

Open

HomesGH opened this issue Jul 30, 2024 · 0 comments

Contributor

HomesGH commented Jul 30, 2024

Describe the bug

Different results (e.g. Epot or p) are computed depending on the number of ranks used.
OpenMP does not have any effect on the result (as expected).

To Reproduce

Steps to reproduce the behavior:

Choose e.g. adios/write/config_mix.xml example
Run simulation for 10 steps with SEQ/PAR and different ranks

The values after 10 steps:

Version	Epot	p
SEQ	-15068.6	-3.72075
PAR 1 rank	-15068.6	-3.72075
PAR 1 rank + OMP	-15068.6	-3.72075
PAR 2 ranks	-15071.4	-3.72071
PAR 2 ranks + OMP	-15071.4	-3.72071
PAR 4 ranks	-15071.4	-3.72071
PAR 8 ranks	-15071.9	-3.72142

Additional context

Maybe this is no proper bug but due to inaccuracies in the communication plus the chaotic system.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment